How a pragmatic, radically simple design produced answers faster than anyone expected, and what trialists took from it afterwards.
The design
Randomisation took under two minutes, data capture rode on records clinicians were already keeping, and the protocol fitted on a few pages. That is what let it recruit at the speed it did.
What it found
Dexamethasone reduced mortality in patients receiving respiratory support, and several widely used candidates were shown to do nothing at all — which was just as valuable.
The lasting lesson
Simplicity in trial design is not a compromise on rigour. It is often what makes rigour achievable.
